Update app.py

app.py

@@ -111,8 +111,8 @@ def fetch_context(db, model, query, logger, template, use_compressor=True):
             compressor.llm_chain.prompt.template = template['llama_rag_template']
             
         elif model_name=='mistral':
-            HF_pipeline_model = wrap_model(model)
             global HF_pipeline_model
+            HF_pipeline_model = wrap_model(model)
             compressor = LLMChainExtractor.from_llm(HF_pipeline_model)
             compressor.llm_chain.prompt.template = template['rag_template']